Skip to content

Conversation

@nealrichardson
Copy link
Collaborator

@nealrichardson nealrichardson commented Nov 7, 2025

Also bumps github action versions and adds latest Connect releases to the matrix.

@nealrichardson nealrichardson changed the title Use with-connect instead of bespoke integration test setup ci: Use with-connect instead of bespoke integration test setup Nov 7, 2025
@github-actions
Copy link

github-actions bot commented Nov 7, 2025

☂️ Python Coverage

current status: ✅

Overall Coverage

Lines Covered Coverage Threshold Status
1998 1884 94% 0% 🟢

New Files

No new covered files...

Modified Files

No covered modified files...

updated for commit: 4c83672 by action🐍

Copy link
Collaborator

@tdstein tdstein left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

In the GitHub Action logs, the status messages from with-connect are showing after the pytests are run. Is that normal? https://github.com/posit-dev/posit-sdk-py/actions/runs/19170009585/job/54799850272#step:5:197

@tdstein
Copy link
Collaborator

tdstein commented Nov 7, 2025

I use the ./integration/Makefile often for local development. There's also a parent target in ./Makefile called make it which forwards to ./integration/Makefile. I'll try refactoring what you have here to fit that scheme.

@nealrichardson
Copy link
Collaborator Author

In the GitHub Action logs, the status messages from with-connect are showing after the pytests are run. Is that normal? https://github.com/posit-dev/posit-sdk-py/actions/runs/19170009585/job/54799850272#step:5:197

Yeah it appears stdout isn't flushing--if you add the timestamps you can see that everything prints at the end. That's not what I had been seeing elsewhere: when I ran on connectapi yesterday it flushed eagerly (https://github.com/posit-dev/connectapi/actions/runs/19150543098/job/54739311800), and that was on the same SHA of with-connect: https://github.com/posit-dev/posit-sdk-py/actions/runs/19170009585/job/54799850272#step:1:42. So maybe it's something with how pytest is running here, maybe the xml reporting?

I think I dropped the -s flag from the pytest invocation because I couldn't tell what it was for and I didn't see it other places, is that relevant?

@tdstein tdstein self-assigned this Nov 7, 2025
…ify CI workflow

Add a new Makefile for integration tests that:
- Dynamically allocates available ports using Python socket binding
- Parses CONNECT_VERSION correctly from Make target patterns
- Supports running tests for individual or all Connect versions
- Includes comprehensive help documentation

Update CI workflow to:
- Extract Connect versions dynamically from the Makefile
- Simplify the integration test matrix setup
- Eliminate duplicate version list maintenance

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <[email protected]>
@tdstein
Copy link
Collaborator

tdstein commented Nov 10, 2025

@nealrichardson - I factored back out into the Makefile and added a script that allocates a random port so that I can run the test suite against multiple versions of Connect in paralle. Might be worth porting that over to with-connect.

@tdstein tdstein requested a review from a team November 10, 2025 16:11
CONNECT_VERSIONS := \
2025.10.0 \
2025.09.1 \
2025.09.0 \
Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

IDK that we need to test 09.0 and 09.1 both, I think we can save the ⚡

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Yeah, I was considering latest patch here too

@nealrichardson
Copy link
Collaborator Author

@nealrichardson - I factored back out into the Makefile and added a script that allocates a random port so that I can run the test suite against multiple versions of Connect in paralle. Might be worth porting that over to with-connect.

posit-dev/with-connect#30

This looks good to me; I can't "approve" because I created the PR, so please merge when you're done with this.

@tdstein tdstein merged commit 326ccdb into main Nov 11, 2025
45 checks passed
@tdstein tdstein deleted the with-connect branch November 11, 2025 21:33
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants